The Generative Art Museum.

What Is Generative Art?

Generative art is an art form where the creation process is driven by autonomous systems, representing a fascinating intersection of creativity and computation. Characterized by the use of algorithms and rule-based systems to produce artworks, the key principles of generative art include:

Algorithmic Creation

The use of algorithms to define the process of creating art, often resulting in complex and intricate patterns.


Systems operate with minimal human intervention, allowing for the generation of unpredictable and unique outcomes.

Complexity and Emergence

Simple rules can lead to complex behaviors and emergent properties, showcasing the depth and unpredictability of the generative process.

Georg Nees, 23-Ecke (poklygon of 23 vertices)
Georg Nees, 23-Ecke (poklygon of 23 vertices), Computer-generated drawing, 1965
A grid of variations of one theme is displayed in this early example of computer-generated art. The drawing was part of the first exhibition of algorithmic art worldwide. It took place in February 1965 in Stuttgart, Germany. (© Courtesy of the sons of Georg Nees)

History of Generative Art

Generative art, defined by its use of algorithms and autonomous systems to create artwork, has a rich and varied history that spans several key periods. This history reflects the evolving relationship between art, mathematics, and technology, culminating in contemporary digital practices. Here is a detailed exploration of the history of generative art.

Early Influences and Precursors

Pre-20th Century

Islamic art: known for its intricate geometric patterns and tessellations, can be seen as an early form of algorithmic art. The use of mathematical principles to create complex designs laid a conceptual foundation for generative processes. Geometric patterns occur in a variety of forms in Islamic art and architecture. These include kilim carpets, Persian girih and Moroccan zellij tilework, muqarnas decorative vaulting, jali pierced stone screens, ceramics, leather, stained glass, woodwork, and metalwork. Interest in Islamic geometric patterns is increasing in the West, both among craftsmen and artists like M. C. Escher in the twentieth century, and among mathematicians and physicists such as Peter J. Lu and Paul Steinhardt.

Darb-e Imam shrine
Girih tiling in the decagonal pattern on a spandrel from the Darb-e Imam shrine.

Renaissance Art: Artists like Leonardo da Vinci used mathematical principles and ratios to create compositions, foreshadowing the algorithmic methods of later generative artists.

Annunciation, Leonardo da Vinci, 1472
Annunciation, Leonardo da Vinci, 1472
A good example of the usage of the golden ratio. Widespread artistic interest in the golden ratio (within Western circles) can be justifiably associated with the publication of Divina proportione in 1509. The book, written by mathematician Luca Pacioli and illustrated by Leonardo da Vinci, was widely lauded for its clear writing and stunning illustrations.

Early 20th Century

Dadaism (1910s-1920s): Artists like Marcel Duchamp and Hans Arp embraced chance and randomness, challenging traditional notions of artistic creation. Duchamp’s use of readymades and Arp’s random collages introduced elements of unpredictability into art.

Portrait of Chess Players
Portrait de joueurs d'echecs (Portrait of Chess Players), Marcel Duchamp, 1911

Surrealism (1920s-1930s): Surrealists like André Breton and Max Ernst explored automatic drawing and the unconscious mind, emphasizing the role of chance and subconscious processes in artistic creation.

Configuration No.6
Configuration No.6, Max Ernst, 1974

Mid-20th Century: The Birth of Generative Art


Ben Laposky created some of the earliest computer-generated art, known as "Oscillons," by manipulating electronic waveforms on an oscilloscope and capturing the results photographically.

Ben Laposky
To achieve these graphic designs, Laposky “manipulated electronic transmissions through the fluorescent surface of an oscilloscope's cathode ray tube (similar to a television tube) and then recorded the abstract patterns using high-speed film, color filters, and special camera lenses”.

John Whitney often referred to as the father of computer graphics, used analog computers to create abstract films, pioneering the use of technology in artistic creation.

Matrix I, John Witney, 1971
Matrix I, John Witney, 1971
video, c., computer-generated
16 mm film, color, sound
black-and-white recording from computer screen, color added through optical printer.

John Cage, an avant-garde composer who used chance operations and the I Ching to create music (Cage's Chance Operations), influencing the generative art ethos of embracing randomness.

Cage's Chance Operations
To achieve these graphic designs, Laposky “manipulated electronic transmissions through the fluorescent surface of an oscilloscope's cathode ray tube (similar to a television tube) and then recorded the abstract patterns using high-speed film, color filters, and special camera lenses”.

In the 1960s and 1970s, the field of generative art continued to evolve as more artists and computer scientists began experimenting with the use of computers for creating art. During this period, many artists were inspired by the concept of "cybernetics," which refers to the study of systems and processes that involve communication and control. These artists saw the potential for computers to be used as a tool for creating complex, self-regulating systems that could produce art that was dynamic and ever-changing.


Frieder Nake a German artist and mathematician, used early computers to create algorithmic drawings, establishing a foundation for future generative art.


Georg Nees, a contemporary of Nake, also used computers to create art, focusing on the interplay between randomness and structure in his algorithmic compositions.

Achsenparalleler Irrweg, Georg Nees, 1964
Achsenparalleler Irrweg, Georg Nees, 1964

A. Michael Noll created computer-generated visual compositions at Bell Labs, exploring the artistic potential of digital technology.

Vertical-horizontal number three, A. Michael Noll, 1962-1963
Vertical-horizontal number three, A. Michael Noll, 1962-1963

Ernest Edmonds, a British artist and pioneer in the field of computer art and its variants, algorithmic art, generative art, interactive art. His work is represented in the Victoria and Albert Museum, as part of the National Archive of Computer-Based Art and Design.

Ernest Edmonds, Nineteen, 1968-69 [photo Jules Lister]
Nineteen, Ernest Edmonds, 1968-69 (photo Jules Lister)


Harold Cohen developed AARON, an artificial intelligence program capable of creating original artworks. AARON’s development marked a significant milestone in the use of AI for generative art.

AARON image created at the Computer Museum, Boston, MA, 1995.
AARON image created at the Computer Museum, Boston, MA, 1995.

Vera Molnar, one of the first women to use computers in art, explored algorithmic and geometric compositions, pushing the boundaries of generative practices.

Hypertransformation 19, Vera Molnar, 1975
Hypertransformation 19, Vera Molnar, 1975

Herbert W. Franke Pioneer of generative art, Herbert Franke used algorithms and computer programs to visualize mathematics in the form of art. His early works revolved around generative photography, and analog computers. He also used the first mainframe computers in the 1950s and 1960s to produce abstract algorithmic art. The Venice Biennale included a screen print from his SQUARE series in 1970. He was the co-founder of Ars Electronica, the largest festival of new media art in the world, in Linz in 1979. With the emergence of NFTs, his writings and artworks have never been more significant.

Tanz der Elektronen, Herbert W. Franke, 1962
Tanz der Elektronen, Herbert W. Franke, 1962

Sol LeWitt: was an American artist linked to various movements, including conceptual art and minimalism. LeWitt came to fame in the late 1960s with his wall drawings and "structures" (a term he preferred to "sculptures") but was prolific in a wide range of media including drawing, printmaking, photography, painting, installation, and artist's books.

Complex Form with Black and White Bands, Sol LeWitt, 1988
Complex Form with Black and White Bands, Sol LeWitt, 1988
Significant art movements, such as conceptualism, minimalism, and generative art are associated with Sol Lewitt. He is most famous for his Wall Drawings, a series of works that highlighted the idea that the concept behind the art is just as essential as the actual artwork itself. A group of drafters accurately captured LeWitt’s vision by following the directions he set for the making of each wall drawing. In essence, LeWitt has created an algorithm that may be used to produce an endless number of variants on the same piece of art. The key difference is that, unlike today’s interconnected computers, LeWitt’s Wall Drawings are created by a team of gallery staff and written in plain English rather than in code.

In the 1980s and 1990s, the field of generative art continued to grow and evolve, and it was influenced by a wide range of factors, including the development of new computer technologies, the rise of the internet, and the proliferation of digital media. During this period, many artists began using computers to generate art that was more interactive and responsive to the viewer.

Late 20th Century: Expansion and Diversification


William Latham and Stephen Todd used evolutionary algorithms to create complex, organic forms. Their work demonstrated the potential of generative systems to produce intricate and lifelike patterns.

Organic Art, William Latham and Stephen Todd
Organic Art, William Latham and Stephen Todd

Manfred Mohr, originally a painter, turned to computer-generated art, creating algorithmic compositions based on the cube and other geometric forms. One notable example of generative art from this period is "Opus 175," a piece created by Mohr in 1969. This piece consisted of a series of abstract, geometric shapes that were generated using a computer program and displayed on a CRT monitor. Mohr's work was significant because it demonstrated the potential for computers to be used for creating complex, abstract visual art, and it influenced a generation of artists who sought to explore the creative possibilities of computational systems.

P-702/A _ enduraChrome, Manfred Mohr
P-702/A _ enduraChrome, Manfred Mohr


Karl Sims used genetic algorithms to evolve virtual creatures and create complex, interactive artworks. His work at the intersection of art and artificial life expanded the possibilities for generative art.

Genetic Images (Screenshot), Karl Sims, 1993
P-702/A _ enduraChrome, Manfred Mohr

John Maeda, a pioneer in computational design, Maeda's work at the MIT Media Lab explored the intersection of art, technology, and design, influencing a generation of digital artists.

Digitalia, John Maeda, 2003
Digitalia, John Maeda, 2003

One notable example of generative art from this period is "The Navigator", a piece created by artist Frieder Weiß in 1993. This piece consisted of a computer program that generated a series of abstract, geometric shapes that were displayed on a screen and constantly changed in response to the movement of the viewer. Weiß's work was significant because it demonstrated the potential for generative art to be interactive and responsive, and it influenced a generation of artists who sought to create immersive, interactive experiences using computational systems.

Early 21st Century: Digital Revolution and Contemporary Generative Art


Processing (Casey Reas and Ben Fry): Reas and Fry developed Processing, an open-source programming language and environment designed for artists. Processing democratized generative art, making it accessible to a broader audience.


Golan Levin created interactive installations and performances using real-time generative systems, engaging audiences in dynamic and participatory experiences.

Stria, Golan Levin, 2002
Stria, Golan Levin, 2002
A system which responds to touch through the development of glowing spots, Stria is a study in infinitesimally slow change. Its interactivity permits its user to create and manipulate a slowly-evolving color field: after establishing a set of initial conditions by adding virtual material to a simulated terrain, the terrain is gradually and procedurally dissolved, in a graphic evolution transpiring on the scale of minutesandhours. Stria exists in several formats, including a touchscreen-based software kiosque, a full-screen software artwork, and a miniatureonline applet.


Generative Adversarial Networks (GANs) The development of GANs by Ian Goodfellow in 2014 revolutionized generative art. Artists like Mario Klingemann used GANs to create highly realistic and innovative artworks, pushing the boundaries of what generative art could achieve.

Refik Anadol used machine learning algorithms to create immersive installations that transformed data into dynamic visual experiences, exploring the intersection of art, technology, and architecture.

Machine Hallucinations, Refik Anadol, 2021
Coral Dreams NFT Collection emerges from Refik Anadol Studio’s ongoing research project on data aesthetics based on collective visual memories of nature and urban environments. For this unique, site-specific data sculpture designed for a specific location in Miami Beach, Anadol and his team collected 300 million nature theme focused data and used 1,742,772 images of coral from publicly available social media platforms and processed them with machine learning classification models. As the machine-mind begins to make its own connections between data points and “hallucinate” about alternative coral shapes and colors, the data universe expands into a latent cosmos in which fluid dynamics becomes the main inspiration of Anadol’s artistic creativity.


NFTs and Blockchain Technology The advent of Non-Fungible Tokens (NFTs) enabled artists to tokenize their generative works, creating verifiable digital ownership and new revenue streams. Platforms like Art Blocks have emerged, allowing artists to mint generative art directly on the blockchain.

Fidenza #598, Tyler Hobbs, 2021
Fidenza #598, Tyler Hobbs, 2021
Fidenza was Hobbs' first long-form generative work, and it remains one of his most notable works to date, comprised of 999 algorithmically generated works, all created by one algorithm without any form of curation. At the time of its release, it was noted for the wide variety of outcomes it produced while still maintaining a cohesive character as a collection.

The history of generative art reflects a continual interplay between creativity, mathematics, and technology. From the early experiments with chance and randomness to contemporary explorations of artificial intelligence and blockchain, generative art has evolved into a dynamic and influential field. As technology continues to advance, generative art will likely keep pushing the boundaries of artistic expression, opening new frontiers for both artists and audiences.

The Emergence of NFTs in Digital Art

The integration of NFTs into the digital art market has been transformative. Artists can tokenize their works, creating digital certificates of ownership that can be bought, sold, and traded. This has opened new revenue streams for artists and established a thriving secondary market. The decentralized nature of blockchain ensures that artists retain a greater share of profits and can even earn royalties on future sales.

Autoglyph #8, 2019
Autoglyph #8, 2019
Autoglyphs are an experiment in generative art, each one unique and created by code running on the Ethereum blockchain. The glyphs were originally created in 2019 by anyone who was willing to donate the creation fee of 0.2Ξ (around $35 at the time) to our chosen charity, 350.org. The creator of each glyph became the first owner of that glyph. After 512 glyphs were created, the generator shut itself off forever and the glyphs are only be available on the secondary market.

Non-Fungible Tokens (NFTs) are unique digital assets verified using blockchain technology. They represent ownership of a specific item or piece of content, and their key features include:

Uniqueness: Each NFT has a unique identifier and metadata that distinguishes it from other tokens.

Ownership and Provenance: Blockchain ensures transparent and verifiable ownership history, preventing counterfeiting and providing clear provenance.

Interoperability: NFTs can be traded across different platforms and marketplaces, increasing their liquidity and reach.

Projects like CryptoPunks, Art Blocks, and Autoglyphs exemplify this trend. Art Blocks, for instance, allows artists to create generative algorithms that produce one-of-a-kind outputs when minted as NFTs. Each piece is generated at the time of purchase, ensuring uniqueness and rarity.

Dead Ringers, Dmitri Cherniak, 2022
Dead Ringers, Dmitri Cherniak, 2022
"Every day in January 2022 I generated a new output from an algorithm, generated a random wallet address, and sent the work to that address. On the final day, January 31st, all the previous Dead Ringers were placed in a 5 by 6 grid and minted to a generated address.
It was almost guaranteed none of the wallets that received Dead Ringers would ever be accessible given there are 16 to power of 40 available addresses. I hoped it would help observers appreciate the vastness of the address space and the underlying security it provides to the network used to distribute the work.
While the distribution of individual Dead Ringers was extremely fair, as it protected against bots and ballot stuffing, it was also extremely unlikely that they would ever be collected. For that reason, observers ascribed value to their scarcity and the odds of receiving one.
The Dead Ringers: Edition flips that notion on its head. Now an edition of January 31, 2022, colloquially known as the Dead Ringers grid, is available to all wallets as a 24-hour timed edition, capped at the size of 16 to the power of 40. That is the total number of possible wallets, and each wallet is entitled to own one if they should wish.
The ERC1155 token points to the same PNG file on Arweave that the original January 31, 2022 referenced, meaning the underlying artwork is exactly the same file as the sought after original.
Each wallet is only eligible to mint one edition, and if a wallet already has one it cannot mint another. There is a 24 hour minting period and the edition size is capped at 16 to the power of 40, the number of all possible Ethereum addresses."